A vehicle-mounted detection system and warning method for hidden dangers of dike emergencies

Abstract

The present invention relates to the technical field of dike detection and early warning, and discloses a vehicle-mounted dike danger and hidden trouble detection system, which includes a vehicle-mounted platform, a ground-penetrating radar, a transient electromagnetic instrument connected to the vehicle-mounted platform, and a slope radar, a three-dimensional laser scanner, a pan-tilt camera, an RTK positioning system, and a multi-functional weather station installed on the top of the vehicle-mounted platform. It also discloses a vehicle-mounted dike danger and hidden trouble early warning method, including a dike internal danger and hidden trouble early warning method and a dike slope deformation early warning method. The detection modes of the present invention include an efficient general survey detection mode, a detailed survey detection mode, and a dike slope deformation monitoring mode, and the early warning methods include a dike internal danger and hidden trouble early warning method and a dike slope deformation early warning method, which can detect and early warn of danger and hidden troubles such as cavities, ant nests, piping channels, local non-compaction, and dike slope deformation at different formation depths inside the dike, and ensure the safety of people's lives and property.

Description

Technical Field

[0001] The present invention relates to the technical field of dike detection and early warning, and specifically to a vehicle-mounted dike danger hidden trouble detection system and an early warning method. Background Technique

[0002] Dikes are the basic construction of water conservancy facilities and play a core role in flood control and fighting. The construction of dikes is related to people's lives, economic benefits and social benefits; modern dikes and dams are mainly divided into two categories: earth-rock dams and concrete dams; earth-rock dams are wide dams built by using soil or stones across large rivers; due to the loose materials, the foundation shakes and water seeps in, reducing the firmness of the dikes; concrete dams are mostly built with concrete, and the main feature is to use their own weight to support the water pressure.

[0003] And during different time periods, there are inevitably hidden troubles in the dikes built with different technologies. Therefore, it is particularly important to conduct a thorough investigation of the distribution of hidden dangers in later-stage dike dangers; therefore, it is necessary to detect the hidden dangers of dikes. The methods for detecting dike hidden dangers can be divided into destructive methods and non-destructive methods. The former includes methods such as pit exploration, trench exploration, well exploration and drilling, etc., and the latter mainly refers to physical detection methods.

[0004] The destructive method has the advantage that the detection results are obvious at a glance, which can intuitively reflect the problems. Without complex analysis and calculation, the hidden dangers and the locations of the hidden danger points existing in the dikes can be known; however, this method is time-consuming and laborious, with low efficiency, and has certain destructiveness and locality. It is difficult for the problems found through individual exploration pits to represent the hidden danger problems of the entire dike, and it is difficult to restore the original appearance of the excavated part, with a certain degree of irreversibility.

[0005] The non-destructive methods mainly include ground penetrating radar method, transient electromagnetic method, high-density electrical method and Rayleigh surface wave method; the focuses of these non-destructive detection methods in various geological detection applications are different. For example, the ground penetrating radar method is suitable for detecting shallow dike hidden dangers; while the transient electromagnetic method is suitable for detecting deep dike hidden dangers; due to the complexity of dike engineering conditions and the limitations of on-site detection conditions, relying solely on a single device cannot meet the requirements of comprehensive and accurate hidden danger detection inside the dikes, and it is difficult to obtain comprehensive, correct and reliable geological information, and it cannot be fully applicable to the detection of dike danger hidden troubles.

[0006] A Chinese patent discloses a time-lapse electrical method detection system for dike hidden troubles (publication number CN108873072A). This patented technology realizes the early warning of hidden troubles - disasters through the comparative analysis of the evolution trends of apparent resistivity amplitude, contour line morphology abnormal range, etc. During measurement, multiple groups of data can be measured with one power supply, thus reducing the number of power supplies and the number of times of running the poles of the measurement electrical sensors, greatly reducing the measurement time and improving the field measurement efficiency. However, its detection means are single, the efficiency of hidden trouble exploration is low, and it is difficult to give early warning. Summary of the Invention

[0007] The object of the present invention is to provide a vehicle-mounted detection system and early warning method for hidden dangers of dike risks, so as to solve the problems put forward in the above-mentioned background technology.

[0008] To achieve the above object, the present invention provides the following technical solutions:

[0009] A vehicle-mounted detection system for hidden dangers of dike risks includes a vehicle-mounted platform, a ground-penetrating radar connected to the vehicle-mounted platform, a transient electromagnetic instrument, and a slope radar, a three-dimensional laser scanner, a pan-tilt camera, an RTK positioning system, and a multi-functional weather station installed on the top of the vehicle-mounted platform. It is characterized in that the ground-penetrating radar consists of a radar host and a transceiver antenna, wherein the transceiver antenna is a medium- and low-frequency shielding combined antenna of 400 MHz and 200 MHz; the transient electromagnetic instrument consists of a transceiver host and a transceiver coil, wherein the transceiver coil adopts a modular multi-turn coil.

[0010] The RTK positioning system is respectively arranged at the end positions of the medium- and low-frequency shielding combined antenna of the ground-penetrating radar and at the center position of the transceiver coil of the transient electromagnetic instrument; the pan-tilt camera is used for large-range scanning and monitoring of the dike monitoring area to obtain video and image information in real time; the multi-functional weather station is used for monitoring meteorological data such as wind speed, wind direction, atmospheric temperature, relative humidity, atmospheric pressure, and rainfall at the dike site.

[0011] As a further scheme of the present invention: the detection system includes three working modes, namely an efficient general survey detection mode, a fine detailed survey detection mode, and a dike slope deformation monitoring mode;

[0012] The working method of the efficient general survey detection mode is as follows:

[0013] The vehicle-mounted platform is slowly driven along the dike top road at a traveling speed of not less than 5 km / h; a narrow-pulse broadband high-frequency electromagnetic wave signal is continuously transmitted underground by the way of continuously dragging the transmitting antenna. When the electromagnetic wave signal propagates inside the dike and encounters the interface of underground media with different electrical properties, reflection, transmission, and refraction occur; the two-way travel time, amplitude, and phase of the reflected electromagnetic wave signal are received by the receiving antenna; the radar host performs data migration processing, echo energy gain, digital filtering, and moving average processing, and after calibration by the RTK positioning system, the depth of the underground media interface is calculated in real time to obtain a two-dimensional radar profile of the shallow layer inside the dike, and then the position and condition of the underground media can be judged according to the spatial position represented by the image, and the spatial position, structure, and distribution of hidden dangers not less than 5 meters deep inside the dike can be detected quickly, comprehensively, and accurately;

[0014] The working method of the fine detailed survey detection mode is as follows:

[0015] Slowly drive the vehicle-mounted platform along the dike road at a traveling speed of not less than 3 km / h; supply a step pulse square wave through the transmitting coil. During the power-off interval, use the receiving coil to observe the secondary induced voltage, and perform strong interference signal rejection, combined filtering, and weak information enhancement processing through the electromagnetic instrument main body to obtain the secondary field attenuation curve and apparent resistivity cross-section map; initially identify the lateral anomaly based on the attenuation characteristics of the secondary field induced electromotive force at different measuring points and different times, and then detect the geological anomaly body according to the calculated apparent resistivity image, and the temporal and spatial distribution of hidden dangers not less than 15 meters deep inside the dike can be judged.

[0016] The working method of the embankment slope deformation monitoring mode is as follows:

[0017] The vehicle-mounted platform is used to achieve all-day, all-weather, and high-precision deformation monitoring of the dike slope area in a fixed-point manner. Through echo signal preprocessing and imaging target recognition, a radar two-dimensional interference image is obtained, and the differential interference phase is calculated. Furthermore, the change trends of deformation amount, deformation speed, and acceleration are analyzed to give early warnings of embankment slope deformation hidden dangers such as slope failure and collapse; at the same time, combined with the three-dimensional model of the dike slope obtained by a three-dimensional laser scanner, the position of the hidden danger area is intuitively corresponded through the radar image and terrain data mapping registration algorithm.

[0018] As a further scheme of the present invention: The detection system can be used for detecting potential dangers and hidden dangers of earth-rock dam slopes and concrete dam slopes; its detection method is as follows:

[0019] For earth-rock dam slopes: First, use the efficient general survey detection mode to detect potential dangers and hidden dangers in the shallow layer inside the dike; then, on the key area route suspected of potential dangers and hidden dangers in the deep layer inside the dike, use the intermediate-frequency shielding antenna of the fine detailed survey detection mode to further clarify the hidden danger situation; finally, use the embankment slope deformation monitoring mode to perform high-precision monitoring of the minute deformation on the surface of the key embankment slopes and complete three-dimensional space modeling; ultimately, the intuitive display of potential dangers and hidden dangers in the surface layer, shallow layer, and deep layer of the earth-rock dam is realized, providing support for on-site detection personnel to investigate and dispose of potential dangers and hidden dangers; in the detection of earth-rock dam slopes, the efficient general survey detection mode uses a 400 MHz intermediate-frequency shielding antenna.

[0020] For concrete dam slopes: First, use the efficient general survey detection mode to detect potential dangers and hidden dangers in the shallow layer inside the dike; then, on the key area route suspected of potential dangers and hidden dangers in the deep layer inside the dike, use the fine detailed survey detection mode to further clarify the hidden danger situation. In the detection of concrete dam slopes, the efficient general survey detection mode uses a 200 MHz intermediate-frequency shielding antenna.

[0021] As a further scheme of the present invention: Under the efficient general survey detection mode, the calculation formula for the depth h of the underground medium interface is as follows:

[0022]

[0023] In the above formula (1), v is the propagation speed of the electromagnetic wave, t is the round-trip time of a single pulse, and x is the horizontal displacement of the ground-penetrating radar;

[0024] When the propagation speed v of the electromagnetic wave in the underground medium is determined, the depth h of the underground reflection interface can be obtained based on the accurately measured round-trip time t of a single pulse and the horizontal displacement x of the ground-penetrating radar.

[0025] As a further solution of the present invention: in the fine detailed detection mode, the calculation method of the secondary induced voltage is as follows:

[0026] S21. By passing a current I0 through the transmitting coil and generating an induced current I1(t) in the target body, according to the electromagnetic induction law, the calculation formula for the induced current I1(t) generated in the target body is:

[0027]

[0028] In the above formula (2), I0 is the current passed through the transmitting coil, M is the mutual inductance between the transmitting coil and the target body, τ = L / R, and R and L are the impedance and inductive reactance of the target body respectively;

[0029] S22. Then, an induced magnetic field H2(t) is excited by the induced current I1(t), and the calculation formula for the induced magnetic field H2(t) is as follows:

[0030] H2(t) = Lv0sinθ / I1(t) (3)

[0031] In the above formula (3), v0 is the speed of the target body cutting the magnetic field, L is the length of the target body cutting the magnetic field, and θ is the angle of the target body cutting the magnetic field;

[0032] S23. A transmitting current I(t) is generated by the transmitting coil, and a secondary magnetic field H2(t) is excited by the transmitting current I(t). The calculation formula for the secondary magnetic field H2(t) is as follows:

[0033] Hi2(t) = H2(t) / I(t) (4)

[0034] In the above formula (4), I(t) is the trapezoidal wave transmitting current; H i2 (t) is the secondary magnetic field of the full-space impact pulse current;

[0035] S24. A secondary induced voltage V t (t) is generated in the receiving coil by the secondary magnetic field H2(t). According to the Faraday electromagnetic induction law, the calculation formula for the secondary induced voltage V t (t) is as follows:

[0036]

[0037]

[0038] In the above formula (5): Sr is the equivalent receiving area of the receiving coil; u(t) is the unit step function; t is the round-trip time of a single pulse; t0 is the start time of turn-off; t1 is the complete turn-off time; μ0 is the magnetic permeability of vacuum; t off is the turn-off time; H s2 (t) is the secondary magnetic field of the positive step current;

[0039] S25. According to the relationship between the secondary magnetic field of the positive step current and the secondary magnetic field H –s2 (t), the calculation formula for the secondary induced voltage V t (t) in the receiving coil is as follows:

[0040]

[0041] Therefore, it can be obtained from the above formula (6) that the secondary induced voltage of the trapezoidal wave turn-off is numerically equal to the algebraic sum of the secondary magnetic field generated by the negative step current at time t0 and the secondary magnetic field generated by the negative step current at time t1; only by calculating the secondary magnetic field H –s2 (t) of the negative step current of the full-space three-dimensional geological model, and then combining the turn-off time of the transmitting current and the equivalent area of the receiving coil, the secondary induced voltage V t (t) in the receiving coil can be obtained.

[0042] As a further solution of the present invention: in the embankment slope deformation monitoring mode, the differential interference phase is calculated as follows:

[0043]

[0044] In the above formula (7), and are the phases during the first and second scan monitoring respectively, is the phase component related to the deformation displacement, and are the phase components caused by atmospheric effects during data acquisition, is the phase component caused by other relevant noise sources, 2kπ is the phase ambiguity coefficient, λ is the wavelength of the electromagnetic wave emitted by the slope radar, and Δd is the deformation displacement of the target during the first and second scan monitoring.

[0045] A vehicle-mounted early warning method for embankment danger and hidden trouble includes an early warning method for internal danger and hidden trouble of the embankment and an early warning method for embankment slope deformation; specifically as follows:

[0046] The hidden danger warning method for the internal danger of the dike aims at the hidden dangers of cavities, ant nests, piping channels, and local non-compaction at different formation depths inside the dike. Based on the detection and imaging results of the ground penetrating radar and transient electromagnetic instrument in the vehicle-mounted dike hidden danger detection system, and through the intelligent geological structure identification technology based on the parallel deep learning network model, the Faster R-CNN and YOLOv3 object detection algorithms in deep learning are applied to the identification of the imaging results of the ground penetrating radar and transient electromagnetic instrument, so as to accurately identify the hidden danger targets of the dike at different formation depths and directly give early warnings.

[0047] The embankment slope deformation warning method aims at the hidden dangers of slope failure and collapse induced by embankment slope deformation. Based on the time series deformation displacement information of the slope radar in the vehicle-mounted dike hidden danger detection system, combined with the three-dimensional model of the embankment slope obtained by the three-dimensional laser scanner, by setting up warning levels and warning parameters, the deformation hidden dangers in the embankment slope monitoring area are graded and warned, and the occurrence time of the danger is judged based on the velocity reciprocal method prediction model.

[0048] As a further scheme of the present invention: the specific steps of the embankment slope deformation warning method are as follows:

[0049] S1. According to the actual needs of the embankment slope monitoring area, four warning levels are set, from high to low: red warning, purple warning, yellow warning, and blue warning; each level corresponds to four warning parameters, and the warning parameters are divided into short-term warning value D s , long-term warning value D l , warning area S ew , and warning duration T ew ;

[0050] S2. Using the time series deformation displacement data diff(x i ,y i ) of the slope radar in the embankment slope accelerated deformation stage and comparing it with the warning deformation value, when both exceed the short-term warning value, long-term warning value, and warning area, that is:

[0051] {diff(x i ,y i )>D s ∩diff(x i ,y i )>D l ∩∑S(x i ,y i )>S ew} (8)

[0052] Then it turns to the judgment of the search warning area. If the warning area is not searched, the warning process ends; if the warning area is searched, it turns to the judgment of the warning state; if it is in the warning state, a node information is added to the current warning and the warning process ends; if it is not in the warning state, it turns to the judgment of the warning duration; if the warning duration is not reached, the warning process ends, and if the warning duration is reached, a warning record and the corresponding node information are generated, and the warning process ends, that is:

[0053] ∑t(x i ,y i )>T ew (9)

[0054] S3. When a red warning is triggered, based on the speed data in the accelerated deformation stage, a short-term prediction before the slope is about to slide is carried out by the speed reciprocal method; assuming that the deformation speed is infinite at the moment when the slope slides, the intersection point of the linear trend line of the reciprocal of the speed and time on the time axis is the slope failure time t f ; the linear fitting relationship between the reciprocal of the speed and time after the starting point is

[0055]

[0056] In the above formula (10), A and B are the deformation constants of the monitored slope, and t SP is the start time of the data used in the prediction model;

[0057] When , the slope failure time can be obtained The above formula (10) can be rewritten as:

[0058]

[0059] Integrating the above formula (11) gives:

[0060]

[0061] In the formula, C is the integral constant term; the relationship between the cumulative deformation displacement value of the slope and time after the start point of data calculation conforms to the logarithmic function of formula (10). Applying the logarithmic function model to fit the data points, the parameters A and B can be obtained, and the predicted slope failure time point can be calculated.

[0062] As a further solution of the present invention: the specific usage methods of the short-term warning value, long-term warning value, warning area and warning duration in the S1 step are as follows:

[0063] Short-term warning value: When performing warning verification at the same level, the short-term warning value is preferentially used for verification, and the warning area is searched for the data within one hour;

[0064] Long-term warning value: If no warning is triggered after entering the warning process based on the short-term warning value, the long-term warning value is verified, and the warning area is searched for data within 24 hours accordingly.

[0065] Warning area: When there are points reaching the warning value in the combined short-term / long-term radar monitoring data, the warning area judgment starts. When consecutive points exceed the warning value and the area of consecutive points reaches the warning area, a warning is triggered.

[0066] Warning duration: A single warning trigger is not sufficient to indicate the danger of the embankment slope area, which may be due to interference, vibration, etc. Consecutive warning triggers should immediately issue a notice for investigation and handling.

[0067] Advantages of the present invention compared with the prior art:

[0068] Through the ground penetrating radar, transient electromagnetic instrument, slope radar, 3D laser scanner, pan-tilt camera, RTK positioning system and multi-functional weather station, the present invention can comprehensively obtain detection results such as the radar profile diagram inside the dike, apparent resistivity cross-section diagram, and deformation displacement on the surface of the dike slope, improving the detection and investigation efficiency of hidden dangers in the dike. Its detection modes include an efficient general survey detection mode, a detailed survey detection mode, and a dike slope deformation monitoring mode, which can overcome the complexity of dike engineering conditions and the limitations of on-site detection conditions. Its warning methods include warning methods for hidden dangers inside the dike and warning methods for dike slope deformation, which can effectively utilize the data results of the detection system to comprehensively judge the hidden dangers of dike risks, and then detect and warn of hidden dangers such as cavities, ant nests, piping channels, local non-compaction, and dike slope deformation at different stratigraphic depths inside the dike, ensuring the life and property safety of the people. Description of the Drawings

[0069] Figure 1 It is a schematic structural diagram of a vehicle-mounted dike risk hidden danger detection system;

[0070] Figure 2 It is a schematic flow diagram of a vehicle-mounted dike risk hidden danger warning method. Detailed Embodiments

[0071] In an embodiment of the present invention, a vehicle-mounted detection system for hidden dangers of dike risks includes a vehicle-mounted platform, a ground penetrating radar connected to the vehicle-mounted platform, a transient electromagnetic instrument, and a slope radar, a three-dimensional laser scanner, a pan-tilt camera, an RTK positioning system, and a multi-functional weather station installed on the top of the vehicle-mounted platform. It is characterized in that the ground penetrating radar consists of a radar host and a transceiver antenna. The transceiver antenna is a medium- and low-frequency shielding combined antenna of 400 MHz and 200 MHz. The shielding antenna has strong anti-interference ability, high resolution, and a theoretical maximum detection depth of up to 10 m, meeting the requirements for detecting shallow hidden dangers inside the dike. The transient electromagnetic instrument consists of a transceiver host and a transceiver coil. The transceiver coil adopts a modular multi-turn coil.

[0072] The RTK positioning system is respectively arranged at the end positions of the medium- and low-frequency shielding combined antenna of the ground penetrating radar and at the center position of the transceiver coil of the transient electromagnetic instrument, so as to accurately mark the spatial positions of the ground penetrating radar and the transient electromagnetic instrument for detecting dike hidden dangers, facilitating the hidden danger investigation and disposal by dike inspection personnel. The pan-tilt camera is used to perform a large-range scanning and monitoring of the dike monitoring area, and real-time video and image information can be obtained. The multi-functional weather station is used to monitor meteorological data such as wind speed, wind direction, atmospheric temperature, relative humidity, atmospheric pressure, and rainfall at the dike site, and timely judge and analyze the monitoring results of the ground penetrating radar, the transient electromagnetic instrument, and the slope radar. For example Figure 1 。

[0073] Preferably, the detection system includes three working modes, namely an efficient general survey detection mode, a fine detailed survey detection mode, and a dike slope deformation monitoring mode.

[0074] The working method of the efficient general survey detection mode is as follows:

[0075] Slowly drive the vehicle-mounted platform along the top road of the levee at a traveling speed of not less than 5 km / h, with a basic line spacing of 2 m. When there are suspected hidden dangers, it can be encrypted to 1 m, and can be widened to 4 m near the toe of the back slope dam; continuously transmit narrow pulse broadband high-frequency electromagnetic wave signals underground by continuously dragging the transmitting antenna. When the electromagnetic wave signals encounter the interface of underground media with electrical differences (such as dielectric constant differences) during propagation inside the levee, reflection, transmission, and refraction occur; receive the two-way travel time, amplitude, and phase of the reflected electromagnetic wave signals through the receiving antenna; perform data migration processing, echo energy gain, digital filtering, and moving average processing through the radar host, and then after calibration by the RTK positioning system, calculate the depth of the underground media interface in real time to obtain the radar two-dimensional profile of the shallow layer inside the levee, and then the position and condition of the underground media can be judged according to the spatial position represented by the image, and the spatial position, structure, and distribution of hidden dangers not less than 5 meters deep inside the levee can be detected quickly, comprehensively, and accurately; specifically: irregular cavities with a depth within 8 m and not less than 0.4 m × 0.3 m × 0.3 m (length × width × height) can be effectively detected; ant nests with a depth within 5 m and not less than 1 m 3 can be effectively detected; pipe erosion channels with a depth within 8 m and a cross-sectional area not less than 0.2 m 2 can be effectively detected;

[0076] The working method of the fine detailed detection mode is as follows:

[0077] Slowly drive the vehicle-mounted platform along the levee road at a traveling speed of not less than 3 km / h, with a basic line spacing of 2 m and a measuring point spacing of 1 m; when there are suspected hidden dangers, the line spacing can be encrypted to 1 m and the measuring point spacing can be encrypted to 0.5 m; supply a step pulse square wave through the transmitting coil, and during the power-off interval, observe the secondary induced voltage using the receiving coil, and perform strong interference signal elimination, combined filtering, and weak information enhancement processing through the electromagnetic instrument main body to obtain the secondary field attenuation curve and apparent resistivity cross-section map; initially identify the lateral anomaly based on the attenuation characteristics of the secondary field induced electromotive force at different measuring points at different times, and then detect the geological anomaly body according to the calculated apparent resistivity image, and the spatio-temporal distribution of hidden dangers not less than 15 meters deep inside the levee can be judged. Specifically:

[0078] irregular cavities with a depth within 20 m and not less than 0.5 m × 0.5 m × 0.3 m (length × width × height) can be effectively detected; ant nests with a depth within 15 m and not less than 1.5 m 3 can be effectively detected; pipe erosion channels with a depth within 20 m and a cross-sectional area not less than 0.2 m 2 can be effectively detected;

[0079] The working method of the levee slope deformation monitoring mode is as follows:

[0080] The vehicle-mounted platform realizes all-day, all-weather, and high-precision deformation monitoring of the levee slope area in a fixed-point manner. Through preprocessing of echo signals and imaging target recognition, a two-dimensional radar interference image is obtained, and the differential interference phase is calculated. Furthermore, the deformation amount, deformation speed, and acceleration change trend are analyzed to warn of potential levee slope deformations such as slope failure and collapse. At the same time, combined with the three-dimensional model of the levee slope obtained by a three-dimensional laser scanner, the position of the potential hazard area is visually corresponded through the radar image and terrain data mapping registration algorithm.

[0081] Preferably, the detection system can be used to detect potential hazards of earth-rock dam slopes and concrete dam slopes; the detection method is as follows:

[0082] For earth-rock dam slopes: First, an efficient general survey detection mode is used to detect potential hazards in the shallow layer inside the levee; then, in the key area route suspected of potential hazards in the deep layer inside the levee, a medium-frequency shielding antenna in the fine detailed survey detection mode is used to further clarify the potential hazards; finally, a levee slope deformation monitoring mode is used to perform high-precision monitoring of the minute deformations on the surface of the key levee slopes and complete three-dimensional space modeling; ultimately, the potential hazards of the surface layer, shallow layer, and deep layer of the earth-rock dam levee are visually displayed, providing support for on-site detection personnel to investigate and dispose of potential hazards. In the detection of earth-rock dam slopes, the efficient general survey detection mode uses a 400 MHz medium-frequency shielding antenna.

[0083] For concrete dam slopes: First, an efficient general survey detection mode is used to detect potential hazards in the shallow layer inside the levee; then, in the key area route suspected of potential hazards in the deep layer inside the levee, the fine detailed survey detection mode is used to further clarify the potential hazards. In the detection of concrete dam slopes, the efficient general survey detection mode uses a 200 MHz medium-frequency shielding antenna.

[0084] Preferably, under the efficient general survey detection mode, the calculation formula for the depth h of the underground medium interface is as follows:

[0085]

[0086] In the above formula (1), v is the electromagnetic wave propagation speed, t is the round-trip time of a single pulse, and x is the horizontal displacement of the ground-penetrating radar.

[0087] When the electromagnetic wave propagation speed v in the underground medium is determined, the depth h of the underground reflection interface can be calculated based on the accurately measured round-trip time t of a single pulse and the horizontal displacement x of the ground-penetrating radar.

[0088] Preferably, under the fine detailed survey detection mode, the calculation method for the secondary induced voltage is as follows:

[0089] S21. Pass a current I0 through the transmitting coil and generate an induced current I1(t) in the target object. According to the law of electromagnetic induction, the calculation formula for the induced current I1(t) generated in the target object is as follows:

[0090]

[0091] In the above formula (2), I0 is the current passed through the transmitting coil, M is the mutual inductance between the transmitting coil and the target object, τ = L / R, and R and L are the impedance and inductive reactance of the target object respectively;

[0092] S22. Then, excite an induced magnetic field H2(t) through the induced current I1(t). The calculation formula for the induced magnetic field H2(t) is as follows:

[0093] H2(t) = Lv0sinθ / I1(t) (3)

[0094] In the above formula (3), v0 is the velocity of the target object cutting the magnetic field, L is the length of the target object cutting the magnetic field, and θ is the angle of the target object cutting the magnetic field;

[0095] S23. Generate a transmitting current I(t) through the transmitting coil, and excite a secondary magnetic field H2(t) through the transmitting current I(t). The calculation formula for the secondary magnetic field H2(t) is as follows:

[0096] H i2 (t) = H2(t) / I(t) (4)

[0097] In the above formula (4), I(t) is the trapezoidal wave transmitting current; H i2 (t) is the secondary magnetic field of the full-space impact pulse current;

[0098] S24. Generate a secondary induced voltage V t (t) in the receiving coil through the secondary magnetic field H2(t). According to the law of Faraday electromagnetic induction, the calculation formula for the secondary induced voltage V t (t) is as follows:

[0099]

[0100] In the above formula (5): Sr is the equivalent receiving area of the receiving coil; u(t) is the unit step function; t is the round-trip time of a single pulse; t0 is the turn-off start time; t1 is the complete turn-off time; μ0 is the magnetic permeability of vacuum; t off is the turn-off time; H s2 (t) is the secondary magnetic field of the positive step current;

[0101] S25. According to the secondary magnetic field of the positive step current and the secondary magnetic field of the negative step current H –s2(t) relationship, the secondary induced voltage V in the receiving coil can be obtained t (t) is calculated as follows:

[0102]

[0103] Therefore, it can be concluded from the above formula (6) that the trapezoidal wave turn-off secondary induced voltage is numerically equal to the algebraic sum of the secondary magnetic fields generated by the negative step current at time t0 and the secondary magnetic field generated by the negative step current at time t1; only the secondary magnetic field H of the negative step current of the full-space three-dimensional geological model needs to be calculated –s2 (t), and then combined with the turn-off time of the transmitting current and the equivalent area of the receiving coil, the secondary induced voltage V in the receiving coil can be obtained t (t).

[0104] Preferably, in the embankment slope deformation monitoring mode, the differential interference phase is calculated as follows:

[0105]

[0106] In the above formula (7), and are the phases during the first and second scan monitoring respectively, is the phase component related to the deformation displacement, and are the phase components caused by the atmospheric effect during the data acquisition period, is the phase component caused by other relevant noise sources, 2kπ is the phase ambiguity coefficient, λ is the wavelength of the electromagnetic wave emitted by the slope radar, and Δd is the deformation displacement of the target body during the first and second scan monitoring.

[0107] A vehicle-mounted embankment danger and hidden danger warning method, including the warning method for internal danger and hidden danger of the embankment and the warning method for embankment slope deformation, so as to effectively utilize the data results of the detection system to comprehensively judge the danger and hidden danger of the embankment; specifically as follows:

[0108] The warning method for internal danger and hidden danger of the embankment aims at the hidden dangers of cavities, ant nests, piping channels, and local non-compaction at different formation depths inside the embankment. Based on the detection and imaging results of the ground penetrating radar and transient electromagnetic instrument in the vehicle-mounted embankment danger and hidden danger detection system, and through the intelligent geological structure recognition technology based on the parallel deep learning network model, the Faster R-CNN and YOL0v3 target detection algorithms in deep learning are applied to the recognition of the imaging results of the ground penetrating radar and transient electromagnetic instrument, and the embankment hidden danger target bodies at different formation depths are accurately recognized and directly warned, as Figure 2 shown;

[0109] The embankment slope deformation early warning method aims at the potential risks of slope sliding and collapse induced by embankment slope deformation. Based on the slope radar time series deformation displacement information in the vehicle-mounted embankment danger and hidden trouble detection system, combined with the three-dimensional model of the embankment slope obtained by the three-dimensional laser scanner, by setting up early warning levels and early warning parameters, it conducts hierarchical early warning on the deformation hidden troubles in the embankment slope monitoring area, and judges the occurrence time of the danger based on the velocity reciprocal method prediction model.

[0110] Preferably, the specific steps of the embankment slope deformation early warning method are as follows:

[0111] S1. According to the actual needs of the embankment slope monitoring area, four early warning levels are set up, from high to low: red early warning, purple early warning, yellow early warning, and blue early warning; each level corresponds to four early warning parameters, and the early warning parameters are divided into short-term early warning value D s , long-term early warning value D l , early warning area S ew , and early warning duration T ew ;

[0112] S2. Use the slope radar time series deformation displacement data diff(x i ,y i ) in the accelerated deformation stage of the embankment slope to compare with the warning deformation value. When both exceed the short-term warning value, long-term warning value, and warning area, that is:

[0113] {diff(x i ,y i )>D s ∩diff(x i ,y i )>D l ∩∑S(x i ,y i )>S ew} (8)

[0114] Then it transfers to the search for the warning area judgment. If the warning area is not found, the warning process ends; if the warning area is found, it transfers to the warning state judgment; if it is in the warning state, a node information is added to the current warning and the warning process ends; if it is not in the warning state, it transfers to the warning duration judgment; if the warning duration is not reached, the warning process ends, and if the warning duration is reached, a warning record and the corresponding node information are generated, and the warning process ends, that is:

[0115] ∑t(x i ,y i )>T ew (9)

[0116] S3. When a red alert is triggered, based on the velocity data in the accelerated deformation stage, a short-term prediction before the slope is about to slide is carried out by the velocity reciprocal method; assuming that the deformation velocity is infinite at the moment when the slope slides, the intersection point of the linear trend line of the velocity reciprocal value and time on the time axis is the slope failure time t f ; The linear fitting relationship between the reciprocal of the velocity and time after the starting point is

[0117]

[0118] In the above formula (10), A and B are the deformation constants of the monitored slope, and t SP is the starting time of the data used in the prediction model;

[0119] When the slope failure time can be obtained The above formula (10) can be rewritten as:

[0120]

[0121] Integrating the above formula (11) gives:

[0122]

[0123] In the formula, C is the integral constant term; the relationship between the cumulative deformation displacement value of the slope and time after the starting point of data calculation conforms to the logarithmic function of formula (10). Applying the logarithmic function model to fit the data points, obtaining the parameters A and B, the predicted slope failure time point can be calculated.

[0124] Preferably, the specific usage methods of the short-term warning value, long-term warning value, warning area, and warning duration in step S1 are as follows:

[0125] Short-term warning value: When performing warning verification at the same level, the short-term warning value is preferably used for verification, and the warning area is searched for data within one hour;

[0126] Long-term warning value: If no warning is triggered after entering the warning process according to the short-term warning value, the long-term warning value is verified, and the warning area is searched for data within 24 hours;

[0127] Warning area: When there are points reaching the warning value in the combined short-term / long-term radar monitoring data, then the warning area judgment starts. When the continuous points exceed the warning value and the area of the continuous points reaches the warning area, a warning is triggered;

[0128] Warning duration: A single warning trigger is not sufficient to indicate the danger of the slope area. The possible reasons may be interference, vibration, etc. Continuous warning triggers should immediately send out a notice for investigation and handling.

[0129] The above are only the preferred specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, making equivalent substitutions or changes, shall be covered by the protection scope of the present invention.

Claims

1. A vehicle-mounted warning method for hidden dangers of dike danger situations, characterized in that, It includes a warning method for hidden dangers inside the levee and a warning method for levee slope deformation; The warning method for hidden dangers inside the levee targets hidden dangers such as cavities, ant nests, piping channels, and locally non-compact areas at different formation depths inside the levee. Based on the detection and imaging results of the ground-penetrating radar and transient electromagnetic instrument in the vehicle-mounted levee hidden danger detection system, and through the intelligent geological structure identification technology based on the parallel deep learning network model, the Faster R-CNN and YOLOv3 object detection algorithms in deep learning are applied to the identification of the imaging results of the ground-penetrating radar and transient electromagnetic instrument, accurately identifying the hidden danger targets inside the levee at different formation depths and directly giving early warnings; The warning method for levee slope deformation targets hidden dangers such as slope slides and collapses induced by levee slope deformation. Based on the time-series deformation displacement information of the slope radar in the vehicle-mounted levee hidden danger detection system and combined with the 3D model of the levee slope obtained by the 3D laser scanner, by setting warning levels and warning parameters, it conducts graded warnings on the deformation hidden dangers in the levee slope monitoring area, and judges the occurrence time of the danger based on the velocity reciprocal method prediction model; The specific steps of the warning method for levee slope deformation are as follows: S1. According to the actual needs of the embankment slope monitoring area, four warning levels are set up, from high to low: red warning, purple warning, yellow warning, and blue warning; each level corresponds to four warning parameters, and the warning parameters are divided into short-term warning value D s , long-term warning value D l , warning area S ew , warning duration T ew ; S2. Use the time series deformation displacement data diff(x i , y i ) of the slope radar in the accelerated deformation stage of the levee slope to compare with the warning deformation value, where (x i , y i ) is the row and column coordinates of the radar image pixel unit of the i-th point. When the short-term warning value, long-term warning value, and warning area are exceeded simultaneously, that is: {diff(x i ,y i )>D s ∩diff(x i ,y i )>D l ∩∑S(x i ,y i )>S ew} (1), Then it transfers to the search for the warning area judgment. If the warning area is not found, continuous data monitoring is carried out; if the warning area is found, it transfers to the warning state judgment; if it is not in the warning state, a node information is added to the current warning, and data monitoring continues; if it is in the warning state, it transfers to the warning duration judgment; if the warning duration is not reached, data monitoring continues; if the warning duration is reached, that is: ∑t(x i ,y i )>T ew (2), A warning record and corresponding node information are generated, and data monitoring continues; S3. When the red warning level is triggered, based on the velocity data in the accelerated deformation stage, short-term prediction before the levee slope slips is carried out through the velocity reciprocal method.

Images